CN107092898A - A kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment - Google Patents
A kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment Download PDFInfo
- Publication number
- CN107092898A CN107092898A CN201710396747.7A CN201710396747A CN107092898A CN 107092898 A CN107092898 A CN 107092898A CN 201710396747 A CN201710396747 A CN 201710396747A CN 107092898 A CN107092898 A CN 107092898A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- signal
- bispectrum
- msup
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1365—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Nitrogen And Oxygen Or Sulfur-Condensed Heterocyclic Ring Systems (AREA)
Abstract
The invention discloses a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment, it is related to some Bitstream signal of wireless communication field, specifically transmitting terminal, signal s (n) is obtained using QPSK mappings；After up-conversion, obtain FM signal p (n), and it is input to output signal Φ (n) in power amplifier, analog signal is obtained after digital-to-analogue conversion is handled, analog signal is sent out and AWGN is added in transmission process, receiving terminal obtains data signal r (n) after analog-to-digital conversion process, and baseband signal is obtained after down coversion, and radio-frequency fingerprint feature is extracted from baseband signal：Bispectrum Energy-Entropy, first order and second order moments；Then, classification based training and test are carried out to radio-frequency fingerprint feature by SVM classifier, obtains category of test result；By the way that category of test result is contrasted with its actual category result, classification accuracy P is obtainedc.The present invention is effectively classified by SVM classifier to radiofrequency signal, and recognition accuracy improves nearly 20% compared to conventional method under low noise.
Description
Technical field
The present invention relates to wireless communication field, specifically a kind of radio frequency based on QPSK signal bispectrum Energy-Entropies and color moment
Fingerprint identification method.
Background technology
In a wireless communication system, the safety problem brought by the opening of wireless network can not be ignored.
Traditional method is based primarily upon the security protocol of cipher mechanism to realize the protection to data integrity and confidentiality,
And the certification of communicating pair identity is provided.But such authentication information is easy to be copied by software by malicious user, deposits
Threatened potential.Even the distinct device of the same model in view of same manufacturer production, in the fabrication process, due to
The differences such as oxidated layer thickness, doping concentration can form individual difference, and these individual differences also can body in the communication signal
It is existing.
The radio-frequency fingerprint of radio communication is extracted and identification technology is exactly to be extracted by analyzing the signal of communication of wireless device
" radio-frequency fingerprint " of equipment, so as to carry out equipment identification.Equipment recognition accuracy is heavily dependent on the choosing of fingerprint characteristic
Select, it is therefore desirable to which research includes device-fingerprint feature in the communication signal, to improve equipment recognition accuracy.
Exist in existing radio-frequency fingerprint identification technology in the case of low signal-to-noise ratio, the recognition accuracy of equipment is not high to ask
Topic；Rectangular integration bispectrum, circulation integral bispectrum and axis integration bispectrum etc. in such as contour integral bispectrum, although with higher-order spectrum
Suppress the feature of Gaussian noise, stick signal amplitude and phase information, while the method for processing is easier and obtains extensively should
With.But, contour integral bispectrum is converted into one-dimensional matrix by selecting different contour integral routes by two-dimentional bispectrum matrix
During, the intrinsic feature of many bispectrum matrixes is have ignored, causes the recognition accuracy under low signal-to-noise ratio unsatisfactory.
The content of the invention
The present invention is directed in existing radio-frequency fingerprint identification technology, and recognition accuracy is not high under the Low SNR existed
The problem of, it is proposed that a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment；
Comprise the following steps that：
Step 1: for some Bitstream signal of ofdm system transmitting terminal, QPSK signals s is obtained using QPSK mappings
(n)；
N=0,1,2 ... N-1, N are QPSK signal s (n) length；
Step 2: by QPSK signal s (n) after up-conversion, obtaining FM signal p (n), and be input to power amplifier
In, output signal Φ (n)；
FM signal p (n) is calculated as follows：
P (n)=s (n) ej2πnfT
Wherein f is the carrier frequency of transmitting terminal,For QPSK signal s (n) sampling interval.
Power amplifier uses Taylor polynomial model, and power amplifier output signal is：
LsFor the exponent number of Taylor polynomial, { alIt is multinomial coefficient.
Step 3: output signal Φ (n) is obtained into analog signal after digital-to-analogue conversion is handled, analog signal is sent out
Come and white Gaussian noise is added in transmission process, receiving terminal obtains data signal r (n) after analog-to-digital conversion process；
R (n)=Φ (n)+υ (n)
υ (n) is white Gaussian noise；
Step 4: data signal r (n) is obtained into baseband signal after down coversion, radio frequency is extracted from baseband signal and is referred to
Line feature；
The fingerprint characteristic of radio frequency includes：Bispectrum Energy-Entropy, first order and second order moments.
Specifically include：
Step 401, data signal r (n) obtained into baseband signal y (n) after down coversion；
Y (n)=r (n) e-j2πnfT
Step 402, the Third-order cumulants c for calculating baseband signal y (n)3y(τ1,τ2), and obtain letter using Third-order cumulants
Number bispectrum B (ω1,ω2)；
Third-order cumulants are calculated as follows：
c3y(τ1,τ2)=E { y* (n) y (n+ τ1)y(n+τ2)}
* complex conjugate, τ are represented1,τ2>=0 represents the difference of time.
To Third-order cumulants c3y(τ1,τ2) carry out two dimensional discrete Fourier transform obtain baseband signal bispectrum B (ω1,ω2)：
ω1,ω2∈ (- π, π] digital angular frequency is represented, resolution ratio isWherein NfftCounted for Fourier transformation；
According to Fourier transformation points construction Nfft*NfftMatrix be baseband signal bispectrum matrix B；In bispectrum matrix BOKThe data of row are B (ω1,ω2) value.
Step 403, utilize baseband signal bispectrum B (ω1,ω2) obtain radio-frequency fingerprint feature --- bispectrum Energy-Entropy；
First, baseband signal bispectrum B (ω are utilized1,ω2) calculate bispectrum energy matrix E midpoints (i, j) bispectrum energy value
Eij；
It is calculated as follows：
Wherein i, j=1,2 ... Nfft。
Then, each point energy sum in bispectrum energy matrix E is calculated, and calculates the energy of point (i, j) in gross energy
Accounting pij；
Each point energy sum is in bispectrum energy matrix E：
Accounting pijIt is expressed as：
Finally, the accounting p using the energy each put in gross energyijCalculate bispectrum Energy-Entropy En；
It is as follows：
Step 404, bispectrum matrix B is converted into ζ-bit gray level image matrixes G；
Representative is rounded downwards；Bm',n'Represent the data at bispectrum matrix B midpoint (m', n')；Gm',n'Represent gray level image square
The data of point (m', n') in battle array G, span is 0~2ζ-1；M', n'=1,2 ... Nfft；
Step 405, the first moment μ and second moment ξ for obtaining gray level image matrix G respectively；
Wherein NB=Nfft×NfftRepresent the sum at gray level image matrix G midpoints.
Step 5: carrying out classification based training and test to radio-frequency fingerprint feature by SVM classifier, category of test knot is obtained
Really.
Specially：
Step 501, the radio-frequency fingerprint feature composition characteristic vector [E that transmitting terminal is extracted to all bit streamsn, μ, ξ], divide
Into sampling feature vectors set D and testing feature vector set D '.
Step 502, with sampling feature vectors set D characteristic vector train SVM classifier；
Step 503, the characteristic vector in testing feature vector set D is input in the SVM classifier trained, obtained
Category of test result.
Step 6: category of test result is contrasted with its actual category result, classification accuracy P is obtainedc。
The advantage of the invention is that：
1), a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment, passes through SVM classifier
Radiofrequency signal can effectively be classified.
2), a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment, in low noise situation
Under recognition accuracy improve nearly 20% compared to traditional contour integral bispectrum feature.
Brief description of the drawings
Fig. 1 be radio-frequency fingerprint recognition methods signal of the present invention based on QPSK signal bispectrum Energy-Entropies and color moment send and
Receive schematic diagram；
Fig. 2 is the schematic diagram of the radio-frequency fingerprint recognition methods of the invention based on QPSK signal bispectrum Energy-Entropies and color moment；
Fig. 3 is the flow chart of the radio-frequency fingerprint recognition methods of the invention based on QPSK signal bispectrum Energy-Entropies and color moment；
When Fig. 4 is the SNR of selection 10dB in the embodiment of the present invention, the radio-frequency fingerprint feature of 3 kinds of equipment is in feature space
Perspective view；
Fig. 5 is the relations comparison chart of the classification degree of accuracy and SNR in the embodiment of the present invention.
Specific embodiment
The specific implementation method to the present invention is described in detail below in conjunction with the accompanying drawings.
A kind of radio-frequency fingerprint identification technology (A Radio based on QPSK signal bispectrum Energy-Entropies and color moment
Frequency Fingerprinting identification method based on bispectrum energy
Entropy and color moments of QPSK modulation signal), with reference to bispectrum Energy-Entropy, and bispectrum
Matrix, which is converted into after two dimensional gray digital picture, obtains image first order and second order moments, and composition 3-dimensional fingerprint characteristic carries out equipment knowledge
Not, the radio-frequency fingerprint identification technology applied to qpsk modulation signal, effectively raises the identification of radio-frequency apparatus under low signal-to-noise ratio
Accuracy rate, so as to ensure communication safety.
As shown in figure 1, handling process of the signal from transmitting terminal to receiving terminal is：
Some Bitstream signal of transmitting terminal, QPSK signal s (n) are obtained using QPSK mappings；After up-conversion, obtain
FM signal p (n), and be input in power amplifier, output signal Φ (n)；Power amplifier is the one of transmitting terminal least significant end
Individual element, its nonlinear characteristic is also the primary identity feature for sending equipment；The present invention is from the non-linear of different power amplifiers
Different embody of the difference of feature in the signal that receiving terminal is received is started with, and is selected and is extracted radio-frequency fingerprint feature, carry out equipment
Identification.By the way that output signal Φ (n) is obtained into analog signal after digital-to-analogue conversion is handled, by analog signal send out and
AWGN is added in transmission process, receiving terminal obtains data signal r (n) after analog-to-digital conversion process, is obtained after down coversion
Baseband signal, extracts radio-frequency fingerprint feature from baseband signal；
Then, classification based training and test are carried out to radio-frequency fingerprint feature by SVM classifier, obtains category of test result.
Detailed process is as shown in Fig. 2 by the way that category of test result is contrasted with its actual category result, obtain classification accuracy Pc。
Such as Fig. 3 institutes, comprise the following steps that：
Step 1: for some Bitstream signal of ofdm system transmitting terminal, QPSK signals s is obtained using QPSK mappings
(n)；
N=0,1,2 ... N-1, N are QPSK signal s (n) length；
Step 2: by QPSK signal s (n) after up-conversion, obtaining FM signal p (n), and be input to power amplifier
In, output signal Φ (n)；
FM signal p (n) is calculated as follows：
P (n)=s (n) ej2πnfT
Wherein f is the carrier frequency of transmitting terminal,For QPSK signal s (n) sampling interval.
Power amplifier uses Taylor polynomial model, and power amplifier output signal is：
LsFor the exponent number of Taylor polynomial, { alIt is multinomial coefficient.
Step 3: output signal Φ (n) is obtained into analog signal after digital-to-analogue conversion is handled, analog signal is sent out
Come and white Gaussian noise is added in transmission process, receiving terminal obtains data signal r (n) after analog-to-digital conversion process；
R (n)=Φ (n)+υ (n)
υ (n) is white Gaussian noise；
Step 4: data signal r (n) is obtained into baseband signal after down coversion, radio frequency is extracted from baseband signal and is referred to
Line feature；
The fingerprint characteristic of radio frequency includes：Bispectrum Energy-Entropy, first order and second order moments.
Specifically include：
Step 401, data signal r (n) obtained into baseband signal y (n) after down coversion；
Y (n)=r (n) e-j2πnfT
Step 402, the Third-order cumulants c for calculating baseband signal y (n)3y(τ1,τ2), and obtain letter using Third-order cumulants
Number bispectrum B (ω1,ω2)；
Third-order cumulants are calculated as follows：
c3y(τ1,τ2)=E { y*(n)y(n+τ1)y(n+τ2)}
* complex conjugate, τ are represented1,τ2>=0 represents the difference of time.
To Third-order cumulants c3y(τ1,τ2) carry out two dimensional discrete Fourier transform obtain baseband signal bispectrum B (ω1,ω2)：
ω1,ω2∈ (- π, π] digital angular frequency is represented, resolution ratio isWherein NfftCounted for Fourier transformation；
According to Fourier transformation points construction Nfft*NfftMatrix be baseband signal bispectrum matrix B；In bispectrum matrix BOKThe data of row are B (ω1,ω2) value.
Step 403, utilize baseband signal bispectrum B (ω1,ω2) obtain radio-frequency fingerprint feature --- bispectrum Energy-Entropy；
First, bispectrum B (ω1,ω2) energy matrix represented with E, utilize baseband signal bispectrum B (ω1,ω2) calculate double
The bispectrum energy value E at spectrum energy matrix E midpoints (i, j)ij；
Bispectrum；
It is calculated as follows：
Wherein i, j=1,2 ... Nfft。
Then, each point energy sum in bispectrum energy matrix E is calculated, and calculates the energy of point (i, j) in gross energy
Accounting pij；
Each point energy sum E is in matrix E：
Accounting pijIt is expressed as：
Finally, the accounting pi using the energy each put in gross energyjCalculate bispectrum Energy-Entropy En；
In order to weigh the degree that is evenly distributed of bispectrum, the present invention represents the distribution situation of bispectrum energy using Energy-Entropy.
If energy is uniformly distributed in two dimensional surface, then its energy entropy is maximum.If conversely, energy is concentrated mainly on some regions
It is interior, then its energy entropy is smaller.
It is as follows：
Step 404, bispectrum matrix B is converted into ζ-bit gray level image matrixes G；
Bispectrum is considered as a two dimensional gray digital picture for containing gray-scale intensity information, with ζ-bit gray level images
Exemplified by, bispectrum matrix is converted into by gray level image matrix by following formula：
Representative is rounded downwards；Bm',n'Represent the data at bispectrum matrix B midpoint (m', n')；Gm',n'Represent gray level image square
The data of point (m', n') in battle array G, span is 0~2ζ-1；M', n'=1,2 ... Nfft；
Step 405, the first moment μ and second moment ξ for obtaining gray level image matrix G respectively；
The present invention describes two gradation of image strength informations point by extracting the single order and second order color moment of gray level image
Cloth situation, first moment μ weighs the mean intensity of gray level image, the standard deviation of second moment ξ representative image gray-scale intensities.
Wherein NB=Nfft×NfftRepresent the sum at gray level image matrix G midpoints.
Step 5: carrying out classification based training and test to radio-frequency fingerprint feature by SVM classifier, category of test knot is obtained
Really.
Specially：
Step 501, the radio-frequency fingerprint feature composition characteristic vector [E that transmitting terminal is extracted to all bit streamsn, μ, ξ], divide
Into sampling feature vectors set D and testing feature vector set D '.
Step 502, with sampling feature vectors set D characteristic vector train SVM classifier；
Step 503, the characteristic vector in testing feature vector set D is input in the SVM classifier trained, obtained
Category of test result.
Step 6: category of test result is contrasted with its actual category result, classification accuracy P is obtainedc。
The Energy-Entropy and color moment of bispectrum are combined as the radio-frequency fingerprint feature [E of equipment by the present inventionn,μ,ξ].Power is put
Taylor polynomial exponent number L in big device Taylor polynomial models=3, and take three groups of { alIt is used as 3 kinds of plant capacity amplifier models
Multinomial coefficient.As shown in figure 4, in signal to noise ratio snr=10dB, the throwing of the radio-frequency fingerprint feature of 3 kinds of equipment in feature space
Understood in shadow figure, the radio-frequency fingerprint feature preferably can make a distinction to distinct device.
On the basis of signal bispectrum Energy-Entropy and color moment is extracted, realize that equipment is recognized using SVM classifier, in difference
By training and detection device under SNR, it recognizes situation as shown in figure 5, by figure it can be found that under low signal-to-noise ratio, equipment is correct
Discrimination Pc80% can be reached more than, and with SNR increase, PcConstantly rise, reached more than in 20dB 95% it is correct
Discrimination.
Claims (4)
1. a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment, it is characterised in that specific step
It is rapid as follows：
Step 1: for some Bitstream signal of ofdm system transmitting terminal, QPSK signal s (n) are obtained using QPSK mappings；
N=0,1,2 ... N-1, N are QPSK signal s (n) length；
Step 2: by QPSK signal s (n) after up-conversion, obtaining FM signal p (n), and it is input in power amplifier,
Output signal Φ (n)；
FM signal p (n) is calculated as follows：
P (n)=s (n) ej2πnfT
Wherein f is the carrier frequency of transmitting terminal,For QPSK signal s (n) sampling interval；
Power amplifier output signal is：
<mrow>
<mi>&Phi;</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>L</mi>
<mi>s</mi>
</msub>
</munderover>
<msub>
<mi>a</mi>
<mi>l</mi>
</msub>
<msup>
<mrow>
<mo>&lsqb;</mo>
<mi>p</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mi>l</mi>
</msup>
</mrow>
LsFor the exponent number of Taylor polynomial, { alIt is multinomial coefficient；
Step 3: output signal Φ (n) is obtained into analog signal after digital-to-analogue conversion is handled, analog signal is sent out simultaneously
White Gaussian noise is added in transmission process, receiving terminal obtains data signal r (n) after analog-to-digital conversion process；
R (n)=Φ (n)+υ (n)
υ (n) is white Gaussian noise；
Step 4: data signal r (n) is obtained into baseband signal after down coversion, radio-frequency fingerprint is extracted from baseband signal special
Levy；
The fingerprint characteristic of radio frequency includes：Bispectrum Energy-Entropy, first order and second order moments；
Step 5: carrying out classification based training and test to radio-frequency fingerprint feature by SVM classifier, category of test result is obtained；
Step 6: category of test result is contrasted with its actual category result, classification accuracy P is obtainedc。
2. a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment as claimed in claim 1,
It is characterized in that：Described step two intermediate power amplifier uses Taylor polynomial model.
3. a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment as claimed in claim 1,
It is characterized in that：Described step four is specially：
Step 401, data signal r (n) obtained into baseband signal y (n) after down coversion；
Y (n)=r (n) e-j2πnfT
Step 402, the Third-order cumulants c for calculating baseband signal y (n)3y(τ1,τ2), and obtain signal bispectrum using Third-order cumulants
B(ω1,ω2)；
Third-order cumulants are calculated as follows：c3y(τ1,τ2)=E { y*(n)y(n+τ1)y(n+τ2)}
* complex conjugate, τ are represented1,τ2>=0 represents the difference of time；
To Third-order cumulants c3y(τ1,τ2) carry out two dimensional discrete Fourier transform obtain baseband signal bispectrum B (ω1,ω2)：
<mrow>
<mi>B</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&omega;</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>&omega;</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>&tau;</mi>
<mn>1</mn>
</msub>
<mo>=</mo>
<mo>-</mo>
<mi>&infin;</mi>
</mrow>
<mrow>
<mo>+</mo>
<mi>&infin;</mi>
</mrow>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>&tau;</mi>
<mn>2</mn>
</msub>
<mo>=</mo>
<mo>-</mo>
<mi>&infin;</mi>
</mrow>
<mrow>
<mo>+</mo>
<mi>&infin;</mi>
</mrow>
</munderover>
<msub>
<mi>c</mi>
<mrow>
<mn>3</mn>
<mi>y</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<msub>
<mi>&tau;</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>&tau;</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>j</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&omega;</mi>
<mn>1</mn>
</msub>
<msub>
<mi>&tau;</mi>
<mn>1</mn>
</msub>
<mo>+</mo>
<msub>
<mi>&omega;</mi>
<mn>2</mn>
</msub>
<msub>
<mi>&tau;</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
</msup>
</mrow>
ω1,ω2∈ (- π, π] digital angular frequency is represented, resolution ratio isWherein NfftCounted for Fourier transformation；According to
Fourier transformation points construction Nfft*NfftMatrix be baseband signal bispectrum matrix B；In bispectrum matrix BOKRow
Data be B (ω1,ω2) value；
Step 403, utilize baseband signal bispectrum B (ω1,ω2) obtain radio-frequency fingerprint feature --- bispectrum Energy-Entropy；
First, baseband signal bispectrum B (ω are utilized1,ω2) calculate bispectrum energy matrix E midpoints (i, j) bispectrum energy value Eij；
It is calculated as follows：
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<msubsup>
<mo>&Integral;</mo>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
<mi>&Delta;</mi>
<mi>&omega;</mi>
</mrow>
<mrow>
<mi>i</mi>
<mi>&Delta;</mi>
<mi>&omega;</mi>
</mrow>
</msubsup>
<msubsup>
<mo>&Integral;</mo>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
<mi>&Delta;</mi>
<mi>&omega;</mi>
</mrow>
<mrow>
<mi>j</mi>
<mi>&Delta;</mi>
<mi>&omega;</mi>
</mrow>
</msubsup>
<mi>B</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&omega;</mi>
<mn>1</mn>
</msub>
<mo>,</mo>
<msub>
<mi>&omega;</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<msub>
<mi>d&omega;</mi>
<mn>1</mn>
</msub>
<msub>
<mi>d&omega;</mi>
<mn>2</mn>
</msub>
</mrow>
Wherein i, j=1,2 ... Nfft；
Then, each point energy sum in bispectrum energy matrix E is calculated, and calculates accounting of the energy of point (i, j) in gross energy
pij；
Each point energy sum is in bispectrum energy matrix E：
Accounting pijIt is expressed as：
Finally, the accounting p using the energy each put in gross energyijCalculate bispectrum Energy-Entropy En；
It is as follows：
<mrow>
<msub>
<mi>E</mi>
<mi>n</mi>
</msub>
<mo>=</mo>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mi>log</mi>
<mi> </mi>
<msub>
<mi>p</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
</mrow>
Step 404, bispectrum matrix B is converted into ζ-bit gray level image matrixes G；
Representative is rounded downwards；Bm',n'Represent the data at bispectrum matrix B midpoint (m', n')；Gm',n'Represent gray level image matrix G
In point (m', n') data, span be 0~2ζ-1；M', n'=1,2 ... Nfft；
Step 405, the first moment μ and second moment ξ for obtaining gray level image matrix G respectively；
<mrow>
<mi>&mu;</mi>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msub>
<mi>N</mi>
<mi>B</mi>
</msub>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<msup>
<mi>m</mi>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<msup>
<mi>n</mi>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<msub>
<mi>G</mi>
<mrow>
<msup>
<mi>m</mi>
<mo>&prime;</mo>
</msup>
<mo>,</mo>
<msup>
<mi>n</mi>
<mo>&prime;</mo>
</msup>
</mrow>
</msub>
</mrow>
<mrow>
<mi>&xi;</mi>
<mo>=</mo>
<msup>
<mrow>
<mo>&lsqb;</mo>
<mrow>
<mfrac>
<mn>1</mn>
<msub>
<mi>N</mi>
<mi>B</mi>
</msub>
</mfrac>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<msup>
<mi>m</mi>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<msup>
<mi>n</mi>
<mo>&prime;</mo>
</msup>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>N</mi>
<mrow>
<mi>f</mi>
<mi>f</mi>
<mi>t</mi>
</mrow>
</msub>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mrow>
<msub>
<mi>G</mi>
<mrow>
<msup>
<mi>m</mi>
<mo>&prime;</mo>
</msup>
<mo>,</mo>
<msup>
<mi>n</mi>
<mo>&prime;</mo>
</msup>
</mrow>
</msub>
<mo>-</mo>
<mi>&mu;</mi>
</mrow>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
<mo>&rsqb;</mo>
</mrow>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
</msup>
</mrow>
Wherein NB=Nfft×NfftRepresent the sum at gray level image matrix G midpoints.
4. a kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment as claimed in claim 1,
It is characterized in that：Described step five is specially：
Step 501, the radio-frequency fingerprint feature composition characteristic vector [E that transmitting terminal is extracted to all bit streamsn, μ, ξ], it is divided into sample
Eigen vector set D and testing feature vector set D '；
Step 502, with sampling feature vectors set D characteristic vector train SVM classifier；
Step 503, the characteristic vector in testing feature vector set D is input in the SVM classifier trained, tested
Category result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710396747.7A CN107092898B (en) | 2017-05-31 | 2017-05-31 | QPSK signal bispectrum energy entropy and color moment based radio frequency fingerprint identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710396747.7A CN107092898B (en) | 2017-05-31 | 2017-05-31 | QPSK signal bispectrum energy entropy and color moment based radio frequency fingerprint identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107092898A true CN107092898A (en) | 2017-08-25 |
| CN107092898B CN107092898B (en) | 2020-09-04 |
Family
ID=59638977
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710396747.7A Active CN107092898B (en) | 2017-05-31 | 2017-05-31 | QPSK signal bispectrum energy entropy and color moment based radio frequency fingerprint identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107092898B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109167641A (en) * | 2018-07-28 | 2019-01-08 | 中国人民解放军战略支援部队信息工程大学 | Narrow-band power amplifier method for extracting fingerprint feature and device |
| CN110061947A (en) * | 2019-03-28 | 2019-07-26 | 东南大学 | A kind of anti-multipath jamming radio-frequency fingerprint extracting method based on OFDM lead code |
| CN111191703A (en) * | 2019-12-24 | 2020-05-22 | 北京百卓网络技术有限公司 | Method and system for analyzing wireless local area network flow |
| CN111262606A (en) * | 2020-01-21 | 2020-06-09 | 东南大学 | Method and device for separating channel characteristics and equipment physical fingerprints based on cepstrum |
| CN112689288A (en) * | 2020-12-18 | 2021-04-20 | 桂林电子科技大学 | Radio frequency fingerprint extraction and identification method based on WANN |
| CN114615666A (en) * | 2022-03-11 | 2022-06-10 | 四川大学 | Universal radio frequency fingerprint extraction and identification method for self-adaptive variable interference domain |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080095264A1 (en) * | 2005-05-27 | 2008-04-24 | Huawei Technologies Co., Ltd. | Device and method for pre-distorting a base-band digital signal |
| CN104065597A (en) * | 2014-07-04 | 2014-09-24 | 哈尔滨工程大学 | BPSK/QPSK signal identification method based on wavelet energy distribution entropy |
| CN105184264A (en) * | 2015-09-14 | 2015-12-23 | 杭州电子科技大学 | RF fingerprint feature extraction method based on Wigner diagonal slice spectrum |
-
2017
- 2017-05-31 CN CN201710396747.7A patent/CN107092898B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080095264A1 (en) * | 2005-05-27 | 2008-04-24 | Huawei Technologies Co., Ltd. | Device and method for pre-distorting a base-band digital signal |
| CN104065597A (en) * | 2014-07-04 | 2014-09-24 | 哈尔滨工程大学 | BPSK/QPSK signal identification method based on wavelet energy distribution entropy |
| CN105184264A (en) * | 2015-09-14 | 2015-12-23 | 杭州电子科技大学 | RF fingerprint feature extraction method based on Wigner diagonal slice spectrum |
Non-Patent Citations (3)
| Title |
|---|
| JINGWEN ZHANG ET AL: "Specific Emitter Identification via Hilbert–Huang Transform in Single-Hop and Relaying Scenarios", 《IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY》 * |
| 徐扬: "辐射源信号指纹识别技术", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 蒋鹏: "雷达信号细微特征分析与识别", 《中国博士学位论文数据库 信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109167641A (en) * | 2018-07-28 | 2019-01-08 | 中国人民解放军战略支援部队信息工程大学 | Narrow-band power amplifier method for extracting fingerprint feature and device |
| CN110061947A (en) * | 2019-03-28 | 2019-07-26 | 东南大学 | A kind of anti-multipath jamming radio-frequency fingerprint extracting method based on OFDM lead code |
| CN110061947B (en) * | 2019-03-28 | 2021-09-07 | 东南大学 | Anti-multipath interference radio frequency fingerprint extraction method based on OFDM lead code |
| CN111191703A (en) * | 2019-12-24 | 2020-05-22 | 北京百卓网络技术有限公司 | Method and system for analyzing wireless local area network flow |
| CN111262606A (en) * | 2020-01-21 | 2020-06-09 | 东南大学 | Method and device for separating channel characteristics and equipment physical fingerprints based on cepstrum |
| CN111262606B (en) * | 2020-01-21 | 2022-01-07 | 东南大学 | Method and device for separating channel characteristics and equipment physical fingerprints based on cepstrum |
| CN112689288A (en) * | 2020-12-18 | 2021-04-20 | 桂林电子科技大学 | Radio frequency fingerprint extraction and identification method based on WANN |
| CN114615666A (en) * | 2022-03-11 | 2022-06-10 | 四川大学 | Universal radio frequency fingerprint extraction and identification method for self-adaptive variable interference domain |
| CN114615666B (en) * | 2022-03-11 | 2023-06-13 | 四川大学 | Universal radio frequency fingerprint extraction and identification method for self-adaptive multi-variable interference domain |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107092898B (en) | 2020-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107092898A (en) | A kind of radio-frequency fingerprint recognition methods based on QPSK signal bispectrum Energy-Entropies and color moment | |
| US10985955B2 (en) | Method for automatically identifying modulation mode for digital communication signal | |
| Zhang et al. | An efficient deep learning model for automatic modulation recognition based on parameter estimation and transformation | |
| Peng et al. | Radio frequency fingerprint identification based on slice integration cooperation and heat constellation trace figure | |
| CN108540202A (en) | A kind of satellite communication signals Modulation Mode Recognition method, satellite communication system | |
| CN107147599B (en) | Automatic map domain feature construction method for communication signal modulation recognition | |
| WO2021103206A1 (en) | Machine learning algorithm-based wireless radio frequency device identity recognition method and system | |
| CN107979842A (en) | A kind of mobile phone individual discrimination method based on bispectrum feature and deep learning | |
| CN111866876A (en) | Wireless equipment identification method and device based on frequency offset preprocessing and storage medium | |
| CN112312457A (en) | Communication radiation source individual identification method based on complex depth residual error network | |
| CN111050315B (en) | Wireless transmitter identification method based on multi-core two-way network | |
| CN108881096B (en) | Spatial modulation MQAM base station based on phase decision | |
| Peng et al. | Supervised contrastive learning for RFF identification with limited samples | |
| CN110087233B (en) | Unmanned aerial vehicle identity authentication method based on radio frequency fingerprint | |
| Zeng et al. | Multi-channel attentive feature fusion for radio frequency fingerprinting | |
| CN104868962B (en) | Frequency spectrum detecting method and device based on compressed sensing | |
| CN110944002B (en) | Physical layer authentication method based on exponential average data enhancement | |
| Aminuddin et al. | Securing wireless communication using RF fingerprinting | |
| Zeng et al. | Visibility graph entropy based radiometric feature for physical layer identification | |
| Morehouse et al. | RF device identification using CNN based PUF | |
| CN104392252B (en) | Emitter Recognition and device | |
| CN108494711B (en) | Communication signal map domain feature extraction method based on KL divergence | |
| CN111222430A (en) | Unmanned aerial vehicle identification method and system based on artificial intelligence | |
| CN112488092A (en) | Navigation frequency band signal type identification method and system based on deep neural network | |
| Cheng et al. | Radio frequency emitter identification based on ensemble learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |